One of the key problems in developmental neurobiology is to find the molecular cues by which neurons recognize their proper synaptic partners. Our strategy is to use behavioral screens and electrophysiological testing to identify genes which disrupt the specificity of neural connectivity. We study a small set of eight neurons which mediate the visually driven startle/escape response in Drosophila melanogaster. We have done extensive physiological and anatomical characterization of this circuit. We have individually studied each of the large cells of this circuit by stimulating and recording from them, filling them with dyes and examining their synapses in the EM. The Giant Fiber, a large axon which descends from the brain to the thorax, is the command neuron for this response. Mutation of the Passover gene disrupts specific connections between the GF and the motorneuron and interneuron which are its thoracic synaptic targets. We have cloned the gene and it codes for a transmembrane protein from a new gene family. The gene is expressed in the adult brain in a pair of large cells which are probably the GFs and in thorax in cells that are probably the synaptic targets of the GF. It is expressed throughout adult life, so is probably involved in maintenance of the synapses as well as their development. We propose to attempt to rescue the phenotype by P-element mediated transformation. In an attempt to determine the mechanism of action of the Pas gene we will study the temporal and tissue specific expression of the gene's transcripts and proteins in wild-type and mutant strains. Since it seems that only a few cells express this gene strongly, we will use electrophysiological and molecular double labelling to determine the identity of the cells. Electron microscopy will be used to look at specifically affected synapses. We will develop tissue and organ culture techniques to study the cells behavior in vitro. We will begin to study how the genome controls Pas expression in such a small number of cells. Using lowstringency hybridization, we will examine other species, including humans, to see if Pas is a member of a gene family which codes for a set of homologous molecules. We expect that the molecules used for specifying neural connections in Drosophila, once they are identified, will prove to be of general significance in the development of nervous systems in widely different animal groups. There may be homologous genes that are used in the development of the human brain; if so, this research may improve our understanding of genetically based human mental disabilities. If similar molecules are used by human descending motor control neurons, this research may provide a basis for therapeutic measures after spinal cord trauma.